Sentence-transformers: Getting SSL Error in downloading "distilroberta-base-paraphrase-v1" model embeddings

[rahuliitkgp31]

I am using google collab with PyTorch version 1.7.0+cu101
I am getting an SSL Error when I am trying to download "distilroberta-base-paraphrase-v1" model.

Code
from sentence_transformers import SentenceTransformer
model = SentenceTransformer('distilroberta-base-paraphrase-v1')

Error

SSLError Traceback (most recent call last)
/usr/local/lib/python3.6/dist-packages/urllib3/connectionpool.py in urlopen(self, method, url, body, headers, retries, redirect, assert_same_host, timeout, pool_timeout, release_conn, chunked, body_pos, **response_kw)
599 body=body, headers=headers,
--> 600 chunked=chunked)
601

24 frames
SSLError: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:852)

During handling of the above exception, another exception occurred:

MaxRetryError Traceback (most recent call last)
MaxRetryError: HTTPSConnectionPool(host='public.ukp.informatik.tu-darmstadt.de', port=443): Max retries exceeded with url: /reimers/sentence-transformers/v0.2/distilroberta-base-paraphrase-v1.zip (Caused by SSLError(SSLError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:852)'),))

During handling of the above exception, another exception occurred:

SSLError Traceback (most recent call last)
SSLError: HTTPSConnectionPool(host='public.ukp.informatik.tu-darmstadt.de', port=443): Max retries exceeded with url: /reimers/sentence-transformers/v0.2/distilroberta-base-paraphrase-v1.zip (Caused by SSLError(SSLError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:852)'),))

During handling of the above exception, another exception occurred:

FileNotFoundError Traceback (most recent call last)
/usr/lib/python3.6/shutil.py in rmtree(path, ignore_errors, onerror)
473 # lstat()/open()/fstat() trick.
474 try:
--> 475 orig_st = os.lstat(path)
476 except Exception:
477 onerror(os.lstat, path, sys.exc_info())

FileNotFoundError: [Errno 2] No such file or directory: '/root/.cache/torch/sentence_transformers/sbert.net_models_distilroberta-base-paraphrase-v1'

[shamusfield-expel]

I'm also seeing this, exact same error for SentenceTransformer('bert-base-nli-stsb-mean-tokens')

[nreimers]

You can try to download the models from here and unzip them (and load from local disc then):
https://public.ukp.informatik.tu-darmstadt.de/reimers/sentence-transformers/v0.2/

[rahuliitkgp31]

@nreimers Yea I manually downloaded the model and it worked for me. Thanks

[anu43]

You can try to download the models from here and unzip them (and load from local disc then):
https://public.ukp.informatik.tu-darmstadt.de/reimers/sentence-transformers/v0.2/

I was working with another model on google colab and don't have a fast internet connection to download and upload every time. When do you think that it will be solved?

[nreimers]

Not sure what the issue is with google colab, so sadly cannot say when it will be solved.

But you can also use colab to download files:

!wget https://public.ukp.informatik.tu-darmstadt.de/reimers/sentence-transformers/v0.2/distilbert-base-nli-stsb-mean-tokens.zip

And the unzip with colab

[apostolosT]

I 've solved this issue in colab by doing the following:
! wget --no-check-certificate https://public.ukp.informatik.tu-darmstadt.de/reimers/sentence-transformers/v0.2/bert-large-nli-stsb-mean-tokens.zip
!unzip /content/bert-large-nli-stsb-mean-tokens.zip -d /root/.cache/torch/sentence_transformers/sbert.net_models_bert-large-nli-stsb-mean-tokens

[nreimers]

I temporally switched the model download to a different server (to a mirror server). It should hopefully work.

I am not sure if the mirror server has all models from the main server. If some model is missing, let me know and I can upload it.

It appears that the authority that signed the SSL certificate (DFN) is somehow not recognized by all systems as legitimate.

[pritamdeka]

Hi @nreimers ,
The following models are also giving an error in colab:
distilroberta-base-paraphrase-v1 and xlm-r-distilroberta-base-paraphrase-v1

[tianzongw]

I 've solved this issue in colab by doing the following:
! wget --no-check-certificate https://public.ukp.informatik.tu-darmstadt.de/reimers/sentence-transformers/v0.2/bert-large-nli-stsb-mean-tokens.zip
!unzip /content/bert-large-nli-stsb-mean-tokens.zip -d /root/.cache/torch/sentence_transformers/sbert.net_models_bert-large-nli-stsb-mean-tokens

@apostolosT @nreimers May I ask where to unzip for different models? tried roberta-large-nli-stsb-mean-tokens by simply replacing model names but failed in the unzip part.

checkdir: cannot create extraction directory: /root/.cache/torch/sentence_transformers/sbert.net_models_roberta-large-nli-stsb-mean-tokens
No such file or directory

[nreimers]

@pritamdeka
Thanks, should be fixed now.

You have to ensure that '/root/.cache/torch/sentence_transformers/' exists as path. Maybe you have to create it before you can run unzip

[pritamdeka]

@tianzongw You need to install sentence-transformers first before unzipping the pre-trained models using the method @apostolosT mentioned. Hopefully it would solve your error.

[nreimers]

The issue should be fixed now